Deep learning to extract the meteorological by-catch of wildlife cameras.

Microclimate-proximal climatic variation at scales of metres and minutes-can exacerbate or mitigate the impacts of climate change on biodiversity. However, most microclimate studies are temperature centric, and do not consider meteorological factors such as sunshine, hail and snow. Meanwhile, remote cameras have become a primary tool to monitor wild plants and animals, even at micro-scales, and deep learning tools rapidly convert images into ecological data. However, deep learning applications for wildlife imagery have focused exclusively on living subjects. Here, we identify an overlooked opportunity to extract latent, ecologically relevant meteorological information. We produce an annotated image dataset of micrometeorological conditions across 49 wildlife cameras in South Africa's Maloti-Drakensberg and the Swiss Alps. We train ensemble deep learning models to classify conditions as overcast, sunshine, hail or snow. We achieve 91.7% accuracy on test cameras not seen during training. Furthermore, we show how effective accuracy is raised to 96% by disregarding 14.1% of classifications where ensemble member models did not reach a consensus. For two-class weather classification (overcast vs. sunshine) in a novel location in Svalbard, Norway, we achieve 79.3% accuracy (93.9% consensus accuracy), outperforming a benchmark model from the computer vision literature (75.5% accuracy). Our model rapidly classifies sunshine, snow and hail in almost 2 million unlabelled images. Resulting micrometeorological data illustrated common seasonal patterns of summer hailstorms and autumn snowfalls across mountains in the northern and southern hemispheres. However, daily patterns of sunshine and shade diverged between sites, impacting daily temperature cycles. Crucially, we leverage micrometeorological data to demonstrate that (1) experimental warming using open-top chambers shortens early snow events in autumn, and (2) image-derived sunshine marginally outperforms sensor-derived temperature when predicting bumblebee foraging. These methods generate novel micrometeorological variables in synchrony with biological recordings, enabling new insights from an increasingly global network of wildlife cameras.

Volatile responses of dwarf birch to mimicked insect herbivory and experimental warming at two elevations in Greenlandic tundra.

Plants release a complex blend of volatile organic compounds (VOCs) in response to stressors. VOC emissions vary between contrasting environments and increase with insect herbivory and rising temperatures. However, the joint effects of herbivory and warming on plant VOC emissions are understudied, particularly in high latitudes, which are warming fast and facing increasing herbivore pressure. We assessed the individual and combined effects of chemically mimicked insect herbivory, warming, and elevation on dwarf birch (Betula glandulosa) VOC emissions in high-latitude tundra ecosystems in Narsarsuaq, South Greenland. We hypothesized that VOC emissions and compositions would respond synergistically to warming and herbivory, with the magnitude differing between elevations. Warming increased emissions of green leaf volatiles (GLVs) and isoprene. Herbivory increased the homoterpene, (E)-4,8-dimethyl-1,3,7-nonatriene, emissions, and the response was stronger at high elevation. Warming and herbivory had synergistic effects on GLV emissions. Dwarf birch emitted VOCs at similar rates at both elevations, but the VOC blends differed between elevations. Several herbivory-associated VOC groups did not respond to herbivory. Harsher abiotic conditions at high elevations might not limit VOC emissions from dwarf birch, and high-elevation plants might be better at herbivory defense than assumed. The complexity of VOC responses to experimental warming, elevation, and herbivory are challenging our understanding and predictions of future VOC emissions from dwarf birch-dominated ecosystems.

Strong isolation by distance among local populations of an endangered butterfly species (Euphydryas aurinia).

The marsh fritillary (Euphydryas aurinia) is a critically endangered butterfly species in Denmark known to be particularly vulnerable to habitat fragmentation due to its poor dispersal capacity. We identified and genotyped 318 novel SNP loci across 273 individuals obtained from 10 small and fragmented populations in Denmark using a genotyping-by-sequencing (GBS) approach to investigate its population genetic structure. Our results showed clear genetic substructuring and highly significant population differentiation based on genetic divergence (F ST) among the 10 populations. The populations clustered in three overall clusters, and due to further substructuring among these, it was possible to clearly distinguish six clusters in total. We found highly significant deviations from Hardy-Weinberg equilibrium due to heterozygote deficiency within every population investigated, which indicates substructuring and/or inbreeding (due to mating among closely related individuals). The stringent filtering procedure that we have applied to our genotype quality could have overestimated the heterozygote deficiency and the degree of substructuring of our clusters but is allowing relative comparisons of the genetic parameters among clusters. Genetic divergence increased significantly with geographic distance, suggesting limited gene flow at spatial scales comparable to the dispersal distance of individual butterflies and strong isolation by distance. Altogether, our results clearly indicate that the marsh fritillary populations are genetically isolated. Further, our results highlight that the relevant spatial scale for conservation of rare, low mobile species may be smaller than previously anticipated.

Camera Assisted Roadside Monitoring for Invasive Alien Plant Species Using Deep Learning.

Invasive alien plant species (IAPS) pose a threat to biodiversity as they propagate and outcompete natural vegetation. In this study, a system for monitoring IAPS on the roadside is presented. The system consists of a camera that acquires images at high speed mounted on a vehicle that follows the traffic. Images of seven IAPS (Cytisus scoparius, Heracleum, Lupinus polyphyllus, Pastinaca sativa, Reynoutria, Rosa rugosa, and Solidago) were collected on Danish motorways. Three deep convolutional neural networks for classification (ResNet50V2 and MobileNetV2) and object detection (YOLOv3) were trained and evaluated at different image sizes. The results showed that the performance of the networks varied with the input image size and also the size of the IAPS in the images. Binary classification of IAPS vs. non-IAPS showed an increased performance, compared to the classification of individual IAPS. This study shows that automatic detection and mapping of invasive plants along the roadside is possible at high speeds.

An Automated Light Trap to Monitor Moths (Lepidoptera) Using Computer Vision-Based Tracking and Deep Learning.

Insect monitoring methods are typically very time-consuming and involve substantial investment in species identification following manual trapping in the field. Insect traps are often only serviced weekly, resulting in low temporal resolution of the monitoring data, which hampers the ecological interpretation. This paper presents a portable computer vision system capable of attracting and detecting live insects. More specifically, the paper proposes detection and classification of species by recording images of live individuals attracted to a light trap. An Automated Moth Trap (AMT) with multiple light sources and a camera was designed to attract and monitor live insects during twilight and night hours. A computer vision algorithm referred to as Moth Classification and Counting (MCC), based on deep learning analysis of the captured images, tracked and counted the number of insects and identified moth species. Observations over 48 nights resulted in the capture of more than 250,000 images with an average of 5675 images per night. A customized convolutional neural network was trained on 2000 labeled images of live moths represented by eight different classes, achieving a high validation F1-score of 0.93. The algorithm measured an average classification and tracking F1-score of 0.71 and a tracking detection rate of 0.79. Overall, the proposed computer vision system and algorithm showed promising results as a low-cost solution for non-destructive and automatic monitoring of moths.

Environmental DNA metabarcoding of cow dung reveals taxonomic and functional diversity of invertebrate assemblages.

Insects and other terrestrial invertebrates are declining in species richness and abundance. This includes the invertebrates associated with herbivore dung, which have been negatively affected by grazing abandonment and the progressive loss of large herbivores since the Late Pleistocene. Importantly, traditional monitoring of these invertebrates is time-consuming and requires considerable taxonomic expertise, which is becoming increasingly scarce. In this study, we investigated the potential of environmental DNA (eDNA) metabarcoding of cow dung samples for biomonitoring of dung-associated invertebrates. From eight cowpats we recovered eDNA from 12 orders, 29 families, and at least 54 species of invertebrates (mostly insects), representing several functional groups. Furthermore, species compositions differed between the three sampled habitats of dry grassland, meadow, and forest. These differences were in accordance with the species' ecology; for instance, several species known to be associated with humid conditions or lower temperatures were found only in the forest habitat. We discuss potential caveats of the method, as well as directions for future study and perspectives for implementation in research and monitoring.

Rapid induction of the heat hardening response in an Arctic insect.

The ability to cope with increasing and more variable temperatures, due to predicted climate changes, through plastic and/or evolutionary responses will be crucial for the persistence of Arctic species. Here, we investigate plasticity of heat tolerance of the Greenlandic seed bug Nysius groenlandicus, which inhabits areas with widely fluctuating temperatures. We test the heat tolerance and hardening capacity (plasticity) of N. groenlandicus using both static (heat knock down time, HKDT) and dynamic (critical thermal maximum, CTmax) assays. We find that N. groenlandicus is able to tolerate short-term exposure to temperatures up to almost 50°C and that it can quickly increase heat resistance following heat hardening. Furthermore, we find that this hardening response is reversible within hours after hardening. These findings contrast with common observations from temperate and tropical insects and suggest high thermal plasticity in some Arctic insects which enables them to cope with extreme temperature variability in their habitats.

Recovery based on plot experiments is a poor predictor of landscape-level population impacts of agricultural pesticides.

Current European Union regulatory risk assessment allows application of pesticides provided that recovery of nontarget arthropods in-crop occurs within a year. Despite the long-established theory of source-sink dynamics, risk assessment ignores depletion of surrounding populations and typical field trials are restricted to plot-scale experiments. In the present study, the authors used agent-based modeling of 2 contrasting invertebrates, a spider and a beetle, to assess how the area of pesticide application and environmental half-life affect the assessment of recovery at the plot scale and impact the population at the landscape scale. Small-scale plot experiments were simulated for pesticides with different application rates and environmental half-lives. The same pesticides were then evaluated at the landscape scale (10 km × 10 km) assuming continuous year-on-year usage. The authors' results show that recovery time estimated from plot experiments is a poor indicator of long-term population impact at the landscape level and that the spatial scale of pesticide application strongly determines population-level impact. This raises serious doubts as to the utility of plot-recovery experiments in pesticide regulatory risk assessment for population-level protection. Predictions from the model are supported by empirical evidence from a series of studies carried out in the decade starting in 1988. The issues raised then can now be addressed using simulation. Prediction of impacts at landscape scales should be more widely used in assessing the risks posed by environmental stressors.

The effects of phenological mismatches on demography.

Climate change is altering the phenology of species across the world, but what are the consequences of these phenological changes for the demography and population dynamics of species? Time-sensitive relationships, such as migration, breeding and predation, may be disrupted or altered, which may in turn alter the rates of reproduction and survival, leading some populations to decline and others to increase in abundance. However, finding evidence for disrupted relationships, or lack thereof, and their demographic effects, is difficult because the necessary detailed observational data are rare. Moreover, we do not know how sensitive species will generally be to phenological mismatches when they occur. Existing long-term studies provide preliminary data for analysing the phenology and demography of species in several locations. In many instances, though, observational protocols may need to be optimized to characterize timing-based multi-trophic interactions. As a basis for future research, we outline some of the key questions and approaches to improving our understanding of the relationships among phenology, demography and climate in a multi-trophic context. There are many challenges associated with this line of research, not the least of which is the need for detailed, long-term data on many organisms in a single system. However, we identify key questions that can be addressed with data that already exist and propose approaches that could guide future research.

Climate change and sexual size dimorphism in an Arctic spider.

Climate change is advancing the onset of the growing season and this is happening at a particularly fast rate in the High Arctic. However, in most species the relative fitness implications for males and females remain elusive. Here, we present data on 10 successive cohorts of the wolf spider Pardosa glacialis from Zackenberg in High-Arctic, northeast Greenland. We found marked inter-annual variation in adult body size (carapace width) and this variation was greater in females than in males. Earlier snowmelt during both years of its biennial maturation resulted in larger adult body sizes and a skew towards positive sexual size dimorphism (females bigger than males). These results illustrate the pervasive influence of climate on key life-history traits and indicate that male and female responses to climate should be investigated separately whenever possible.